Infarct Analysis to Explain Outcomes of Perinatal Stroke

Abstract

Neonatal arterial ischemic stroke (NAIS) is the most common form of stroke in children, potentially leading to long-term neurologic sequelae, including cerebral palsy (CP), epilepsy, cognitive impairment, and behavioral problems. Predicting outcomes of NAIS is a challenge. One child with NAIS may have significant impairment, whereas another with a similar infarct may be almost asymptomatic. A tool that could predict a child's outcome based on infarct characteristics (location and volume) would be a powerful aid in the clinic. Multiple studies have shown that infarct characteristics can predict motor impairment such as CP; however, most of these did not use the most sophisticated and sensitive imaging techniques. [1][2][3] The correlation between language and cognitive skills has been less clear.In this issue of Neurology ® , Mackay et al. 4 used voxel-based lesion-symptom mapping (VLSM) to quantitatively test at the voxel level whether specific infarct locations correlate with functional outcomes. This method has been extensively used in adult studies, 5 but has rarely been used to analyze outcomes in children with NAIS. [5][6][7] Mackay and colleagues studied 2 Australian and Swiss cohorts whose data were prospectively collected. MRI scans were obtained from infants with acute NAIS, and outcomes were assessed at a median age of 2.1 years with the Pediatric Stroke Outcome Measure (PSOM) and clinical examination. It is important to note that the authors used a neonatal brain-specific imaging atlas developed at the Royal Children's Hospital, Melbourne, Australia, along with a neonatal brain tissue segmentation tool. The authors tested the relationship with outcome for voxels that were infarcted in at least 5% of patients. They used the statistical analytical package that accompanied their VLSM analytical program, SPM12, and used the Family-Wide Error correction of that program.